Pulmonary tuberculosis (TB) is characterized by oxidative stress and lung tissue destruction. The relationship between these distinct processes and the implications for TB diagnosis and clinical disease staging are poorly understood. Our previously published studies in this area have focused on the anti-oxidant enzyme heme oxygenase-1 (HO-1) in both human and experimental TB infection. Our collaborative studies in India, Brazil and the NIH clinical center identified HO-1 as a biomarker that distinguishes patients with active TB disease from those with latent or successfully treated Mtb infection. Currently our work on human TB is focused on a collaboration with colleagues at the University of Cape Town (funded in part by a joint US-South Africa UO-1 grant with Rob Wilkinson and Alan Sher as co-PIs). An important component of this project is to extend our analysis of host inflammatory markers to a larger, better characterized treatment cohort. At a broader level, the project seeks to compare immune responses (both myeloid and lymphoid) in different tissue sites (e.g. blood, lung, heart) in different forms and stages of tuberculosis. Although the project was initially delayed because of problems in obtaining approval for our clinical protocol, at the time of this report our Cape Town colleagues have successfully enrolled and collected samples from over 240 TB, HIV-TB infected and control subjects. Once the enrollment goal is met (hopefully later in the year) the immunologic analysis of patient samples will commence. In the meantime this collaboration has provided important data supporting results from murine and macaque studies from Dan Barber's lab (see his report) showing that CD153 expression defines a subset of CD4+T cells associated with control of Mtb infection (Sallin et al., in press). In previously published experiments we showed that administration of tin protoporphyrin IX (SnPPIX), a well-characterized HO-1 enzymatic inhibitor, to mice during acute Mtb infection results in substantial reductions in pulmonary bacterial loads comparable to that achieved following conventional antibiotic therapy. We additionally demonstrated that the efficacy of SnPPIX treatment in reducing bacterial burden is dependent on the presence of host T lymphocytes and the cytokine IFN-gamma. In experiments performed during the report period we investigated both the cellular source of HO-1 expression in infected mouse lung as well as the role of T cells and IFN-gamma in the HO-1 response. This analysis revealed that at 3wks post-infection Ly6C+ myeloid cells (presumably inflammatory monocytes) constitute over 80% of the HO1 positive population in the lung parenchyma and by western blot analysis of sorted cells expressed substantially higher levels of HO-1 than Ly6C- myeloid cells, neutrophils and alveolar macrophages sorted from the same infected lung tissue. In animals lacking T cells or IFN-gamma, we observed a delay in the accumulation of Ly6C+ myeloid cells that was replaced by a massive recruitment of neutrophils which became the predominant cell expressing HO-1 although at lower levels. Both in vitro studies and experiments with chimeric mice indicated that neither IFN-gamma or NOS2 (a major effector stimulated by IFN-gamma) are required for the cell intrinsic induction of HO-1 by MTb infection. These data argue that the role for the Th1 response in the activity of the HO-1 inhibitor SnPPX in controlling bacterial growth instead stems from direct co-operation between the infection limiting functions of the drug and Th1 immunity rather than from a requirement for the immune response in HO-1 induction. Our main hypothesis for how inhibition of HO-1 leads to enhanced control of Mtb infection is through reduced generation from heme of free iron an important nutrient needed for intracellular bacterial growth. Consistent with this concept, we showed last year in a related line of experiments that M. tuberculosis infection induces an increase in serum levels of hepcidin, a peptide hormone that regulates the expression of ferroportin, a protein that exports iron from the cytoplasm to the extracellular compartment. Indeed, we also observed that MTb infection is associated with a simultaneous decrease in the expression of ferroportin by pulmonary myeloid cells. Neverthess, in experiments performed during the year we were unsuccessful in linking these responses in vivo using a drug reported to block hepcidin-ferroportin interaction in vitro. We are currently pursuing other strategies to test the role of hepcidin induction and ferroportin reduction in promoting Mtb infection. IL-1 beta is a cytokine which we and other have shown to play a protective role in Mtb infection. At the same time the cytokine can trigger inflammatory pathology when produced in excess. Lysosomal cathepsins are also associated with tissue inflammation and in work published this year (Amaral et al, 2018) we asked whether they play a role in the inflammasome mediate activation of pro-IL-1 into mature bioactive IL-1 beta. Pharmacological inhibition of cathepsin activity resulted in a substantial reduction of both mature IL-1 production and caspase-1 activation in infected macrophages. Moreover, cathepsin inhibition abolished the interaction between NLRP3 and ASC, measured by immunofluorescence imaging in H37Rv-infected macrophages, demonstrating a critical role of the enzyme in NLRP3-inflammasome activation. These findings support a mechanistic link between cathepsin production and IL-1 beta maturation in the inflammatory response to Mtb. Necrotic cell death during Mycobacterium tuberculosis (Mtb) infection is considered host detrimental since it facilitates mycobacterial spread. As introduced in last years report we have been examining the role in tuberculosis of ferroptosis, a type of regulated necrosis induced by the accumulation of free iron and toxic lipid peroxides. We observed that Mtb-induced macrophage necrosis is associated with reduced levels of glutathione and glutathione peroxidase-4 (Gpx4) along with increased free iron, mitochondrial superoxide and lipid peroxidation, all of which are important hallmarks of ferroptosis. Moreover, necrotic cell death in Mtb-infected macrophage cultures was suppressed by ferrostatin-1 (Fer-1), a major ferroptosis inhibitor. Additional experiments revealed that pulmonary necrosis in acutely infected mice is associated with reduced glutathione and Gpx4 as well as increased lipid peroxidation and is likewise inhibited by Fer-1 treatment. Importantly, Fer-1-treated infected animals also exhibited marked 1 log or greater reductions in bacterial load. Together, these findings (manuscript in preparation) implicate ferroptosis as a major mechanism of necrosis in Mtb infection and as a potential target for host directed therapy in tuberculosis.
Showing the most recent 10 out of 26 publications
